Cleanup

Author: AndreasArvidsson

packages/common/src/util/regex.ts

@@ -30,16 +30,20 @@ function makeCache<T, U>(func: (arg: T) => U) {
 export const rightAnchored = makeCache(_rightAnchored);
 export const leftAnchored = makeCache(_leftAnchored);
 
+export function matchAllIterator(text: string, regex: RegExp) {
+  // Reset the regex to start at the beginning of string, in case the regex has
+  // been used before.
+  // See https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/RegExp/exec#finding_successive_matches
+  regex.lastIndex = 0;
+  return text.matchAll(regex);
+}
+
 export function matchAll<T>(
   text: string,
   regex: RegExp,
   mapfn: (v: RegExpMatchArray, k: number) => T,
 ) {
-  // Reset the regex to start at the beginning of string, in case the regex has
-  // been used before.
-  // See https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/RegExp/exec#finding_successive_matches
-  regex.lastIndex = 0;
-  return Array.from(text.matchAll(regex), mapfn);
+  return Array.from(matchAllIterator(text, regex), mapfn);
 }
 
 export function testRegex(regex: RegExp, text: string): boolean {

packages/cursorless-engine/src/processTargets/modifiers/scopeHandlers/SurroundingPairScopeHandler/RangeIterator.ts

@@ -2,29 +2,26 @@ import type { Range } from "@cursorless/common";
 
 /**
  * An iterator that allows for efficient lookup of ranges that contain a search item.
+ * The items must be sorted in document order.
  */
 export class RangeIterator<T extends { range: Range }> {
   private index = 0;
 
-  constructor(
-    public items: T[],
-    sortItems = false,
-  ) {
-    if (sortItems) {
-      this.items.sort((a, b) => a.range.start.compareTo(b.range.start));
-    }
-  }
+  /**
+   * @param items The items to iterate over. Must be sorted in document order.
+   */
+  constructor(public items: T[]) {}
 
   contains(searchItem: Range): boolean {
     return this.advance(searchItem);
   }
 
   getContaining(searchItem: Range): T | undefined {
-    if (!this.advance(searchItem)) {
-      return undefined;
+    if (this.advance(searchItem)) {
+      return this.items[this.index];
     }
 
-    return this.items[this.index];
+    return undefined;
   }
 
   private advance(searchItem: Range): boolean {

packages/cursorless-engine/src/processTargets/modifiers/scopeHandlers/SurroundingPairScopeHandler/getDelimiterOccurrences.ts

@@ -1,5 +1,6 @@
-import { matchAll, Range, type TextDocument } from "@cursorless/common";
+import { matchAllIterator, Range, type TextDocument } from "@cursorless/common";
 import type { LanguageDefinition } from "../../../../languages/LanguageDefinition";
+import type { QueryCapture } from "../../../../languages/TreeSitterQuery/QueryCapture";
 import { getDelimiterRegex } from "./getDelimiterRegex";
 import { RangeIterator } from "./RangeIterator";
 import type { DelimiterOccurrence, IndividualDelimiter } from "./types";
@@ -21,20 +22,14 @@ export function getDelimiterOccurrences(
     return [];
   }
 
-  const delimiterRegex = getDelimiterRegex(individualDelimiters);
-
   const captures = languageDefinition?.getMultipleCaptures(document, [
     "disqualifyDelimiter",
     "textFragment",
   ]);
-  const disqualifyDelimitersIterator = new RangeIterator(
-    captures?.disqualifyDelimiter ?? [],
-    true, // Sort items
-  );
-  const textFragmentsIterator = new RangeIterator(
-    captures?.textFragment ?? [],
-    true,
+  const disqualifyDelimitersIterator = createRangeIterator(
+    captures?.disqualifyDelimiter,
   );
+  const textFragmentsIterator = createRangeIterator(captures?.textFragment);
 
   const delimiterTextToDelimiterInfoMap = Object.fromEntries(
     individualDelimiters.map((individualDelimiter) => [
@@ -52,20 +47,36 @@ export function getDelimiterOccurrences(
     return textFragmentsIterator.getContaining(range)?.range;
   };
 
-  const text = document.getText();
+  const matchIterator = matchAllIterator(
+    document.getText(),
+    getDelimiterRegex(individualDelimiters),
+  );
+
+  const results: DelimiterOccurrence[] = [];
 
-  return matchAll(text, delimiterRegex, (match): DelimiterOccurrence => {
+  for (const match of matchIterator) {
     const text = match[0];
     const range = new Range(
       document.positionAt(match.index!),
       document.positionAt(match.index! + text.length),
     );
 
-    return {
-      delimiterInfo: delimiterTextToDelimiterInfoMap[text],
-      isDisqualified: isDisqualified(range),
-      textFragmentRange: getTextFragmentRange(range),
-      range,
-    };
-  });
+    if (!isDisqualified(range)) {
+      results.push({
+        delimiterInfo: delimiterTextToDelimiterInfoMap[text],
+        textFragmentRange: getTextFragmentRange(range),
+        range,
+      });
+    }
+  }
+
+  return results;
+}
+
+function createRangeIterator(
+  captures: QueryCapture[] | undefined,
+): RangeIterator<QueryCapture> {
+  const items = captures ?? [];
+  items.sort((a, b) => a.range.start.compareTo(b.range.start));
+  return new RangeIterator(items);
 }

packages/cursorless-engine/src/processTargets/modifiers/scopeHandlers/SurroundingPairScopeHandler/getSurroundingPairOccurrences.ts

@@ -25,15 +25,10 @@ export function getSurroundingPairOccurrences(
   for (const occurrence of delimiterOccurrences) {
     const {
       delimiterInfo: { delimiterName, side, isSingleLine },
-      isDisqualified,
       textFragmentRange,
       range,
     } = occurrence;
 
-    if (isDisqualified) {
-      continue;
-    }
-
     let openingDelimiters = openingDelimiterOccurrences.get(delimiterName);
 
     if (isSingleLine) {

packages/cursorless-engine/src/processTargets/modifiers/scopeHandlers/SurroundingPairScopeHandler/types.ts

@@ -48,13 +48,6 @@ export interface DelimiterOccurrence {
    */
   range: Range;
 
-  /**
-   * If `true`, this delimiter is disqualified from being considered as a
-   * surrounding pair delimiter, because it has been tagged as such based on a
-   * parse tree query.
-   */
-  isDisqualified: boolean;
-
   /**
    * If the delimiter is part of a text fragment, eg a string or comment, this
    * will be the range of the text fragment.